The latest Elvis extravaganza is the work of London-based artist Marco Brambilla: a four-minute, hyper-detailed, image-dense video starring the King in his different incarnations—from young army officer to swaggering movie star to bloated has-been—as well as the dream-city of Vegas itself, which somewhat similarly evolves from a small desert oasis into the neon epicenter of debacle-spectacle. In keeping with its themes of celebrity-image proliferation and saturation, King Size (2023) is Brambilla’s biggest video collage yet, using 1,000 looped video clips from different movies, including all 33 Elvis films.
“I’m using the language of excess,” says Brambilla. “I wanted it to be a spectacle in the tradition of Hollywood, Busby Berkeley and Irwin Allen. It’s really over the top.”
So is the technology. Made for a screen that spans 160,000 sq. ft and curves over 180 degrees, the video has an extremely high, 16K resolution. It was commissioned by the artistic director Willie Williams for the band U2 to play during their concerts at the Sphere (29 September-16 December), with other artists such as John Gerrard providing videos as well. King Size will play, appropriately enough, during the group’s 1991 hit “Even Better than the Real Thing”; Brambilla then plans to show the work at his Berlin gallery, Michael Fuchs Galerie, using an Elvis-inspired soundtrack.
The artist has worked with CGI (computer-generated imagery) for his earlier “digital-psychedelic” productions like Heaven’s Gate (2022). This time, he also used artificial intelligence (AI) to create likenesses of Elvis and the outlandish Vegas settings, with some memorable results. One image, created with the Midjourney AI image generator, shows Elvis rising from stacks of gold coins on a gaming table—his body made of gold. Another resulted from feeding the Stable Diffusion generator the prompt: “What would Elvis look like if he were sculpted by the artist who made the Statue of Liberty?”
Brambilla turned to AI instead of human researchers or video editors this time because, he says, he was on deadline to complete this project quickly, within four months. “What I found is that it was very good at sketching, making conceptual sketches and hybrid images. It often comes up with things that are very magical,” he says. “I was able to upload my library of 12,000 to 15,000 video clips into an AI engine and then train the AI to find images I wanted to use,” whether it was an image of a limo in front of a hotel or a neon sign in front of a sunset.